Online markets are a major priority for the CMA. Their development is fundamental to cross-border and global trade. Online information and search tools help consumers identify and compare competitive offers. Online platforms, devices and mobile payments facilitate e-commerce. All of this improves consumer choice and empowerment while supporting new entrants, particularly innovators who challenge traditional business models and markets, thereby increasing competition and economic growth.
Unsurprisingly, most of the competition law discussion on such issues concerns anticompetitive aspects. In doing so, it tends to focus on the restraint in question and its impact on the interaction between firms. I would suggest that these issues be considered more from the consumer’s perspective. To consumers, online and offline are not really separate. Many shopping journeys do take place purely along one of these channels, but the line is increasingly getting blurred. For example, some shoppers start searching in the real world, then go online to compare and make the purchase; some do the reverse; and some start online, then use a shop but make the final purchase online, or in fact compare products online while inside a store. The online and offline options are therefore merging for the consumer, and consequently for suppliers, who are increasingly using both channels to provide their services.
More online activity also increases the scope for businesses to be more responsive to consumer demand, leading to better products and services, more personalisation of offers, better targeting of spending, and a more efficient market overall. However, while online commerce brings significant benefits for consumers, businesses and markets, there are also risks that we need to understand better. In doing so, it is critical that we look at the facts of each market, rather than just the route to market, and have a view on how a potentially anticompetitive restraint might affect the market environment as a whole, rather than considering just the particular form of restraint that is used.
The potential risks of online commerce are of two types.
Consumer risks: alongside the huge benefits that the Internet provides for consumers in terms of improved information and choice, new sources of consumer detriment are emerging. For example, fake reviews, undisclosed commercial blogging, complex tariff structures, misleading or skewed search results, drip pricing, concealed recurring payments, and complex bundling can all harm consumers’ interests and distort competition. Problems may be even worse in markets characterised by information asymmetries, consumer biases (including inertia) or vulnerability. It is therefore important for competition authorities to act quickly and comprehensively, recognising that the online world is international, while ensuring that consumers do not lose trust in online markets.
Competition risks: while the Internet facilitates increased competition, existing firms in the market may seek to prevent or restrict competition from new entrants and business models. For example, ‘bricks and mortar’ firms may seek to restrict online business models. They may try to protect their margins or claim that they need to restrict such competition in order to compete on non-price elements such as service and presentation in store, or they may use efficiency-based arguments based on limiting free-riding, improving service, or protecting brand signalling of quality—all of which can be legitimate concerns. However, in such cases it is necessary to address the difficult question of balancing efficiency, exclusivity and exclusion.
Whatever the practice, once it is identified as harmful, the key question is how to act quickly and comprehensively enough to protect consumers. Indeed, there are very different schools of thought regarding vertical restraints and price and non-price abuses of a dominant position—whether in the online world or not. Competition authorities from different market traditions and with different policies, priorities and responsibilities tend to differ in approach. But the cross-border and even global reach of some online practices, and the speed with which they proliferate, make timely and comprehensive enforcement across jurisdictions—where competition or consumers are negatively affected by those practices—critical. This makes it even more important that we work together and learn from each other to get it right in our own particular cases.
To contribute to that process, I discuss below some selected practices along with some questions, and conclude with a suggestion on how one might approach such practices.
Price floors: Internet minimum advertised pricing (IMAP)
IMAP acts to set a price floor on online sales. Like most price restrictions of this type, the potential concerns are obvious to competition authorities. This is why there is such a suspicion of these practices, and why the burden tends to shift relatively swiftly to the firm that is imposing IMAP to try to justify its conduct, usually based on the efficiency arguments relating to preventing free-riding and thereby ensuring quality and service.
The CMA has concerns about the use of IMAP and other vertical restraints that prohibit the advertising of any prices online, or ban online sales outright. In cases prioritised due to their impact on vulnerable consumers and on online commerce, as well as their precedential and deterrent value, the OFT found infringements by manufacturers of mobility scooters and a number of their respective retailers. Specifically, one firm had banned online sales and the advertising of prices online, while another prohibited online advertising of prices below its own recommended retail prices.
As restrictions liable to reduce consumer choice go, such practices could not be more obvious. They reduce price transparency and significantly increase search costs for consumers, which is of particular concern where the Internet is important for searching and consumers face high search costs due to their circumstances. In this case, by definition these types of scooter tend to be sold to consumers who find it difficult to visit bricks-and-mortar stores, which are few and far between. Most obviously, however, IMAP restrictions soften intra-brand competition, reduce discounting, and can mean higher prices for consumers.
Online cases can raise difficult issues, however: while the OFT found that, in these cases, the restrictions had as their natural consequence the prevention, restriction or distortion of competition, through our casework and our own research/analysis we are still working to understand the range of scenarios in which such practices will be harmful.
Price ceilings and platforms
Restraints that set price ceilings are more tricky for an authority to assess, because they may appear on the surface to be a boon for consumers, allowing them to focus on the non-price aspects of the product that they want, on the understanding that they will get a great deal. This would initially seem all the more helpful when implemented through an online platform that quickly allows providers to make offers that buyers can compare and choose from easily.
In some cases, however, this can lead to a softening of competition, with the ceiling becoming a floor. In some cases, firms are able to offer the ‘best price guarantee’ only because they ‘know’ that no one will undercut them. Or perhaps the mechanism itself is acting to dissuade entry of lower-priced products, since that important distinguishing element has effectively been removed. This can operate to soften price/commission competition on the platforms, and any resulting increased costs of sales may be passed through by third-party sellers to consumers.
Price parity and price relativity agreements have come up in a number of antitrust cases. In relation to Amazon’s price parity requirement for third-party sellers listing products on Amazon Marketplace, for example, parallel Article 101 cases were run by the OFT and Germany’s Bundeskartellamt. This demonstrates the ongoing concerns that competition authorities have about retail most-favoured-nation clauses (MFNs), and their potential to soften price competition between platforms and exclude new entrants—thereby reducing consumer choice. The OFT was ‘minded to close’ the case after Amazon indicated that it would essentially remove the parity clause from its contracts. It is notable that Amazon did not remove parity only in the countries where the competition authorities were taking cases, but across the entire EU.
Preventing selective distributors from selling through platforms
Manufacturers of branded goods naturally want to protect their image, which they might argue may be undermined by sales through a third-party platform. However, in considering restrictions on platform sales, there are different approaches in different jurisdictions, with some authorities doubting both the need for such brand protection and any efficiencies arising from such platform bans. Some find it difficult to see platform bans as indispensable, and press firms for other means of achieving the same aim.
As a member of the European Competition Network, the CMA works closely with other authorities to understand the approaches they are taking in these areas. The CMA applies its prioritisation principles in a manner which considers the work of other European authorities, and we have seen directly how some national investigations may lead to pan-European changes to the manner in which companies’ distribution systems are set up—particularly in relation to online sales. We are also carrying out research to understand more about both the effect of certain restraints in an online world and the enforcement approaches of our colleagues in other jurisdictions.
Steering amid differences
Learning from each other is one thing, but suggesting that authorities adopt a similar approach to practices about which there are widely different opinions is not likely to be successful. For example, different authorities also have very different approaches towards resale price maintenance (RPM). While the CMA treats allegations of RPM—including in the online context—very seriously as an ‘object’ infringement (because RPM has such a high potential for restricting competition and increasing prices for consumers), in some jurisdictions (e.g. the USA) there are well-known changes from prohibition per se to an effects-based analysis. Other authorities look for evidence of an agreement, coercion and monitoring, while others do not require proof of all of these elements.
The differences in investigative methods and approaches are relevant even in an online context, because, although online markets span borders, and many firms’ distribution practices are regional or even global, they can still vary by country due to levels of e-commerce, market structure, language issues and local preferences or biases. In addition, different authorities have different priorities, investigative methods and economic priors about such conduct.
Each case will therefore depend on the facts and on the evidence put to the relevant national competition authority in the context of its market and the relevant enforcement priorities. What I think is most important, however, is the evidence in each particular case.
New wine in old bottles? Object or effect?
In addition to the practices discussed above, there are many others—i.e. access to platforms, online-targeted advertising, and personalised pricing—and controversial differences of opinion about search rankings.
One pervasive question that naturally arises in such assessments is how online practices should be reviewed. Are they themselves so novel that traditional analytical approaches seem outdated and inappropriate? I don’t think so. First, and to a very large extent, anticompetitive practices online are just new wine in old bottles. There is nothing particularly novel about pricing restraints, for example. If an authority or court analyses such restraints as object infringements when they occur in traditional markets then this will also be appropriate when firms practise them online. The same rationale applies for practices that have tended to be subject to effects-based analysis. Naturally, while each case needs to be assessed on its facts and in its relevant context, there is no reason to change approach just because the practice goes online.
What about truly novel online practices, however? Is the object approach inappropriate in such cases? Should an authority be required to prove that the practice is, on balance, harmful using a full effects-based analysis?
My answer is ‘no’, on both counts. I think this sort of stricture on authorities would be too much, and would be particularly inappropriate in fast-moving online markets. Depending on the restraint, an object approach can still be the most appropriate enforcement approach. There are some limits, however, just as there are in the offline world.
If any practice is to be investigated as an object infringement then this should occur only when the practice is anticompetitive by its very nature. Such practices are those that have such a high potential for harm to competition that actual effects do not need to be proven. They include certain types of behaviour (such as price fixing and market sharing) that experience has shown to be likely to produce negative effects on the market.
Even in the case of a novel practice, about which authorities may have no experience themselves, an object approach can be used if they can show how it is inherently anticompetitive. In such situations, case decisions should explain in what respect the restraint in question reveals the requisite harm in order to be characterised as a restriction by object. This may even be relatively brief, depending on the restraint and the facts, but it should exist. This allows the law to evolve and keep up with changing business practices, and provides companies with an explanation of why these practices are being characterised in this manner, thus ensuring that authorities are less likely to be accused of ‘stretching the object box’.
Back to the consumer: exercising choice
When assessing potential harmful practices in the online world, it is important to consider the impact on consumer choice. Some might say that focusing on choice cannot make much of a difference. After all, doesn’t every restraint of competition—even every merger—restrict choice? I think there is something deeper than that going on, and competition authorities should take more note of impacts on consumer choice. Price is one factor, and a big one, but we all know that competition takes place over a range of factors. However, even more than that, it is critical to note that the price, quality, or any other factor of the product by itself is not determinative in driving competition. It is the process of consumers exercising their choice that is the fundamental driver. Competition involves an interaction not only between rivals, but also among these rivals and the consumer. In a well-functioning market, active consumers exert pressure on firms to improve their product. Reasonably well-informed, confident, largely rational consumers exercising choice activates competition among firms.
Some of this thinking is, of course, already embedded in substitutability analysis and market definition, but further focus on whether and how consumer choice is exercised, particularly in analysing online activity, will allow authorities to better appreciate the actual mechanism at play in competition. Do consumers have a choice? How informed is it? Can they exercise their choice? How do they exercise it? How does a restraint affect consumer choice?
I should stress that this is not a choice standard replacing an efficiency standard, but being built more into—and informing—an efficiency standard. Through an increased focus on, and understanding of, the process of the exercise of consumer choice, enforcers and courts can better appreciate the dynamic between providers and customers, and better choose whether and how to intervene.
Doing so would not make us think all that differently, but—like consumers using the Internet—we would be markedly better informed. Note, however, that my suggestion is not about choice itself, but about its exercise. I do not view every reduction of choice as a problem, nor every increase in choice as a solution. I am just suggesting that in assessing alleged restraints we make sure we are aware of what really affects consumers’ interaction in the market. I hope that a more informed process of analysis will lead to even better outcomes.
I would like more direct recognition of how practices are affecting how consumers exercise choice, particularly online. If e-commerce is opening markets, improving and facilitating choice, and thereby empowering consumers, then we need to make sure that this is not restricted. If there are significant and artificial restrictions on consumers’ ability to choose effectively, we need to analyse them closely. Incumbents that fear competition and seek to protect their existing business models at the expense of consumer choice should tread carefully.